Evaluation of inflammation
markers in urine using ELISA
Anna Tallbo, Bachelor Thesis
Biomedicinprogrammet, Uppsala Universitet Tutors
Jenny Rubin Kristin Blom Per Venge
2 Abstract
This project focuses on two proteins released from eosinophil and neutrophil granulocytes. These two proteins are Eosinophil Protein X (EPX) from eosinophils and Human Neutrophil Lipocalin (HNL) from neutrophils. The proteins can both be found in urine. Generally when handling urine samples precipitates can form in the samples, binding large amounts of the protein that is going to be quantified. If you try to eliminate these precipitates by centrifugation or filtration, you also lose large amounts of protein and the quantification will not be correct. The main objective of this project was to evaluate a urine-precipitation-dissolvation (UPD)-buffer from MAIIA-Diagnostics, Uppsala Sweden, as a useful tool to dissolve these precipitates and correctly quantify and measure the levels of EPX and HNL in urine samples. Urine samples were collected from individuals with no regard to their current health status. The quantification method used in this project was two different sandwich ELISAs, one for HNL and one for EPX. Results showed that the UPD-buffer increased the protein levels in some samples, as there seems to be individual variations in how much precipitate is formed during handling of the urine samples. It was also clear that the use of the UPD-buffer never lowered the amount of protein in the samples, compared to untreated samples. Results showed the need for well established and standardized methods when analyzing these kinds of proteins, and the need to collect reference values from healthy individuals for comparison. Sammanfattning
3
Introduction
There are a large variety of leukocytes in the human immune system, divided into the mononuclear leukocytes and the polymorphonuclear leukocytes. The polymorphonuclear leukocytes are called granulocytes and include the eosinophil, the neutrophil and the basophil. The activity of these different cells depends on the nature of the inflammation or the infection (Venge 2004). For example, neutrophils are active during bacterial infections, while an increase in eosinophil numbers can be seen in patients with allergic diseases such as asthma (Bochner 2000). Due to the specificity of the cells, and their differing activity during various inflammatory responses, proteins released from these granulocytes have been proposed, and are currently used, as biomarkers for different inflammatory conditions such as asthma, allergic dermatitis or bacterial infections. This project focuses mainly on the neutrophil and the eosinophil granulocyte.
The neutrophil is an important granulocyte of the innate immune system. It comprises 60-70% of the circulating leukocytes in normal healthy individuals. The neutrophil has a nuclei consisting of 2-5 lobes, and is about 12 µm in diameter. The reserve of neutrophils in the human body is divided between a circulating and a marginal pool of neutrophils (Witko-Sarsat et al. 2000). The circulating pool is mostly present in the larger blood vessels, and the marginal pool consists of stored neutrophils in more narrow capillaries, such as pulmonary capillaries (Peters 1998). The neutrophil can therefore have an immune surveillance function, but can also rapidly transition between a circulating state and an extremely adhesive state when activated by inflammatory signals from inflamed tissues. The neutrophil is the first cell to migrate and arrive at the site of an inflammation, arriving hours before lymphocytes or monocytes (Witko-Sarsat et al. 2000). The adhesion and extravasation of neutrophils to inflamed tissues involves the combination of different chemokines and cytokines.
Their ability to migrate through endothelial layers and reach the site of inflammation starts with a rolling process, where the neutrophils are anchored to the endothelium of blood vessels by interacting with L-, and P-selectins (Sundd et al. 2011). The adhesion of neutrophils to P- and E-selectins will cause the neutrophil to adhere loosely to the endothelium, a process called tethering. Once adhered to the epithelial surface, the neutrophil will bind more firmly to L-selectins, causing the neutrophil to start a rolling-step which binds it more firmly to the endothelial cells (Li et al. 1998). The rolling step also involves E-selectins (Lawrence et al. 1993). Once firmly adhered to the endothelium, the neutrophils can start to transmigrate through the endothelium layer. This occurs at the borders of the endothelial cells, where the junctions between the cells are disorganized. Once the neutrophils have transmigrated into the damaged tissue, the migration is guided by chemotaxis through chemoattractants such as interleukin-8 produced by damaged tissues or formyl peptides originating from bacteria (Witko-Sarsat et al. 2000).
4 differing between an acute viral or an acute bacterial infection. It has been seen that measurements of HNL in blood or urine could be a method with higher specificity and sensitivity to determine whether an infection is viral or bacterial, compared to the traditional measurements of C-Reactive Protein (CRP) in blood. The amount of HNL is increased during bacterial infections, but hardly ever increased during a viral infection (Xu et al. 1995)
The eosinophil is another important granulocyte of the innate immune system. The eosinophil has a bilobed nucleus and is about the same size as the neutrophil although much fewer in numbers. It comprises 2-4% of the circulating leukocytes in healthy individuals, although the lifetime of the eosinophil in the circulation is only a couple of hours. The eosinophil is produced in the bone marrow under the influence of three important cytokines: interleukin (IL)-1, IL-3 and Granulocyte-macrophage colony-stimulating factor (GM-CSF) (Hogan et al. 2008). The release of the eosinophil from the bone marrow to the circulation is triggered by IL-5 (Collins et al. 1995). Normally the dominant population of eosinophils is present in the gastrointestinal tract, as well as the thymus, mammary glands and the uterus where they play a part in local homeostasis and inflammatory control. The recruitment of the eosinophils to these sites is guided by eotaxin-1, and regulated by eosinophil chemokine receptor CCR3 (Humbles et al. 2002).
In T-helper 2 (Th2)-type immune responses eosinophils can be recruited to inflammatory sites by the help of many different cytokines, the most important one of these being IL-5. IL-5 has also been shown to be an essential signal for the recruitment of eosinophils from the bone marrow into the lung after exposure to allergens, causing an allergic (or Th-2) reaction locally (Sanderson 1992). Eosinophils are activated in inflammation reactions by stimulating factors such as uric acid from injured tissue, or by prostaglandin (PGD)-2 from mast cells. They are also activated by proteolytic enzymes, mostly serine proteases, that are synthesized in different microbes, allergens and fungi (Miike et al. 2003). Therefore, the levels of circulating eosinophils are elevated in allergic conditions, helminth infections and also in fungal infections.
The eosinophil, like the neutrophil, contains large numbers of granules which in turn contain different inflammatory proteins and cytokines (Hogan et al. 2008). Many of these pro-inflammatory cytokines can potently induce inflammation responses in various allergic conditions such as allergic dermatitis or asthma (Tischendorf et al. 2000). Degranulation occurs when eosinophils are activated by the previously mentioned mechanisms (such as uric acid or PGD-2) but also eotaxin-1 and CCR3 (Bystrom et al. 2011). Eosinophils contain specific granules that are packed with (among others) four large basic and cytotoxic proteins (Kita 2011):
• Major Basic Protein (MBP) is a highly cytotoxic protein which binds and disturbs bilipid-layers, or cell membranes. By doing this, MBP is toxic to helminths, tumour cells and other human cells. • Eosinophil Peroxidase (EPO) can generate reactive oxidants and free radicals in activated eosinophils - chemicals that kill a large variety of microorganisms. EPO can also bind to microorganisms and potentiate their killing from phagocytes. • Eosinophil Cationic Protein (ECP) is a member of the RNase A-superfamily. ECP has RNase-activity which is needed for its neurotoxic and antiviral effects.
• Eosinophil Protein X (EPX) is also a member of the RNase A-superfamily. EPX, just as ECP, have RNase-activity although EPX has about 100x more RNase-activity than ECP.
5 Both HNL and EPX are currently used as biomarkers for inflammatory diseases; HNL for differing between an acute viral infection and an acute bacterial infection, EPX as an inflammation marker for different allergic inflammatory diseases such as asthma. Both HNL and EPX are excreted in the urine. EPX is not only excreted in patients, but also in the urine of healthy individuals (Morioka et al. 2000). Another important aspect of the EPX-excretion is that amounts of EPX excreted are dependant of circadian rhythms, with levels being highest in the first morning urine sample (Wolthers et al. 2003). The amount of HNL occurring normally in urine is highly individual, depending on a lot of factors, such as current infections or life-style.
The possibility of using urine samples for measuring levels of inflammation markers is an attractive concept. Urine samples are readily available, non-invasive and can be collected even from very small children (Gore et al. 2003). Although, there are problems concerning urine samples; when collecting and storing urine samples the samples are often frozen or cooled, and then thawed again. When urine samples are frozen and then thawed precipitates and protein complexes can form. The protein that is to be quantified can be bound in these precipitates, and if the precipitates are removed by centrifugation or filtration large amounts of protein can also be removed together with the precipitates which will lead to a faulty quantification and a misleading analysis of the protein-concentration in the sample. As a rule these precipitates cannot be seen with the naked eye, and an estimate of the amount of precipitates formed is very difficult.
MAIIA-diagnostics in Uppsala, Sweden, has created a urine-precipitation-dissolvation buffer (UPD-buffer) which is currently being used successfully for erythropoietin-analysis of urine samples in doping tests (Lonnberg et al. 2008). It is believed that this UPD-buffer could also be used to dissolve precipitates formed in urine samples collected for analysis of HNL and EPX. The aim of this project was to see whether the UPD-buffer from MAIIA Diagnostics could be useful in the quantification and analysis of HNL and EPX in urine samples. This was established by comparing the results between frozen and thawed samples and fresh samples, which were either treated with the UPD-buffer or left untreated. The quantification method used for this project was two different sandwich ELISAs, one for HNL and one for EPX.
Materials and Methods
Sandwich ELISA6 Gathering and grouping of samples
To get materials for this project urine samples were collected from 5 volunteers at the Department of Clinical Chemistry at the Academic Hospital in Uppsala, Sweden. These individuals were randomly selected with no regard to health status. Urine samples that were to be tested for EPX were collected from the first morning urine, while the samples tested for HNL were collected with no regard to circadian rhythms. The urine samples were either frozen in aliquots, or kept in a refrigerator for a maximum of 5 days, the refrigerated urine samples being considered as fresh. During the experiments the urine samples were then divided into three different groups for each protein:
- Frozen aliquots treated with UPD-buffer - Frozen aliquots untreated with UPD-buffer - Fresh aliquots treated with UPD-buffer
It is important to note that the samples were not treated with UPD-buffer before freezing of the aliquots, but was added after thawing of the samples just before the experiment.
The layout of one experiment was that different groups were compared to each other, such as frozen aliquots that had been treated with UPD-buffer were compared to frozen aliquots that had not been treated with UPD-buffer. The different samples were typically run in four wells or more in the sandwich ELISA described below, with each of these ELISAs being run typically three times. Preparation of samples
At the beginning of each experiment the samples that were to be treated with UPD-buffer (kindly provided by MAIIA-diagnostics, Uppsala, Sweden) were diluted 9 parts sample to 1 part UPD buffer - typically 90 µl of sample with 10 µl of UPD-buffer - and incubated for 10 minutes in room temperature. The untreated samples were centrifuged for 5 min at 1000 rpm, and the supernatant used. After preparation the samples were diluted with an assay diluent (see the section Buffers and antibodies below) to the appropriate concentrations. The samples that were to be analyzed for EPX were typically diluted 50x, while the samples that were to be analyzed for HNL were diluted 80x. Sandwich ELISA, EPX and HNL
A sandwich ELISA was performed, using pre-coated plates (Mercodia, Uppsala, Sweden). The samples were added to the plate and incubated for 60 min in RT. The plate was then washed using a Anthos Fluido plate-washing machine with a washing buffer consisting of PBS with 0,05% Tween. The primary antibody solution (see the section Buffers and antibodies below) was added to the plate using a multipipette and incubated for 60 min in RT. The plate was then washed using the same
washing buffer as mentioned above.
7 Buffers and antibodies
Assay diluent
HNL: 0,1% Tween, 0,05% CTAB (Merck), 0,2% BSA (Sigma), 0,01M EDTA (tritriplex III) (Merck) and 0,02% NaN3 (Merck).
EPX: 0,05% Tween, 0,1% CTAB (Merck), 0,2% BSA (Sigma), 0,01M EDTA (tritriplex III) (Merck) and 0,02% NaN3 (Merck).
The assay diluents were filtered through a 0,2 µm filter and kept refrigerated between experiments. Primary antibody
HNL: anti-HNL monoclonal mouse IgG 764 (Diagnostics Development), biotinylated. Diluted 4,4 μl antibody in 11 ml of filtered PBS for a full plate.
EPX: anti-EPX monoclonal mouse IgG 616 (Diagnostics Development).Diluted 81,5 μl antibody in 11 ml of filtered PBS for a full plate.
Secondary antibody/Conjugate
HNL: Avidin-HRP (Horseradish Peroxidase) together with a conjugate buffer StabilZyme HRP (SurModics, Eden Praire, USA). Diluted 5,5l of conjugate in 11 ml of conjugate buffer for a full plate.
EPX: Polyclonal Rabbit Anti-Mouse Immunoglobulins/HRP (Dako Denmark, Glostrup, Denmark). Diluted 5,5L in 11 ml of filtered PBS for a full plate.
Substrate
3, 3', 5, 5'-Tetramethyl-benzidine liquid substrate for ELISA (Sigma). Measurements of urinary creatinine
The levels of urinary creatinine were measured in the samples that were to be tested for EPX at the routine lab at the Department of Clinical Chemistry at the Academic Hospital, Uppsala, Sweden. Statistics
All statistical methods were performed using Microsoft Excel 2007.
- Students T-test was used to see differences in individuals when comparing treated/untreated or fresh/frozen samples. p<0.05 was considered significant.
- 95% CI for all groups. - Standard deviations.
8
Results
HNLThe analysis of HNL was divided into two parts. One experiment using frozen and thawed urine samples that were either treated with UPD-buffer or left untreated as a reference. The other experiment compared fresh samples to frozen and thawed samples with both being treated with UPD-buffer. Samples were collected with no regard to circadian rhythms.
The results of the first experiment, using frozen and thawed urine, are shown in the upper part of table 1 and a scatter plot of the results can be seen in figure 1. The samples are numbered 1, 2 and 3. Sample 1 was run 14 times, giving a mean concentration of 40,66 ng/ml for the samples not treated with UPD-buffer and a mean concentration of 44,61 ng/ml for the samples treated with UPD-buffer. Sample 2 was run 12 times giving a mean concentration of 19,66 ng/ml for samples not treated with UPD and a mean concentration of 20,27 ng/ml for samples treated with UPD-buffer. Sample 3 was run 15 times giving a mean concentration of 0,87 ng/ml for samples not treated with UPD and a mean concentration of 0,99 ng/ml for samples treated with UPD-buffer.
Table 1. A summary of the statistics of HNL. Mean indicates the mean concentration of n number of
experiments. The CV% is calculated, as well as a 95% CI for each group and individual. The p-value (p<0.05 is considered significant) for each sample is also indicated as calculated by a student’s t-test.
The mean values for the concentration of HNL in urine ranged from 0,869 ng/ml for sample 3 to 44,61 ng/ml for sample 1. There was a statistically significant difference between the treated samples and the untreated samples in sample 1 (p=0,017) and sample 3 (p=0,004), with values being higher after treatment with the UPD-buffer. There was not a statistically significant difference in sample 2 (p=0,395). Unfortunately there are no reliable reference intervals for the normal mean concentration of HNL in healthy individuals at this time.
The second experiment, comparing fresh samples to frozen samples with all samples treated with UPD-buffer, are shown in the lower part of table 1 and a scatter plot of the experiment can be seen in figure 2. The samples are numbered 4, 5 and 6. Sample 4 was run 12 times, giving a mean concentration of 5,13 ng/ml for the fresh samples treated with UPD-buffer and a mean concentration of 7,04 ng/ml for the frozen samples treated with UPD-buffer. Sample 5 was run 13 times giving a mean concentration of 29,36 ng/ml for the fresh samples treated with UPD-buffer and a mean concentration of 31,66 ng/ml for the frozen samples treated with UPD-buffer. Sample 6 was run 13 times giving a mean concentration of 32,90 ng/ml for the fresh samples treated with
Sample Mean n CV% 95%CI p
1 -UPD 40,66 14 17% 37,09-44,23 0,017 1 +UPD 44,61 14 19% 40,11-49,11 2 -UPD 19,66 12 12% 18,30-21,02 0,395 2 +UPD 20,27 12 18% 18,17-22,36 3 -UPD 0,87 15 7% 0,84-0,90 0,004 3 +UPD 0,99 15 12% 0,92-1,05 4 fresh +UPD 5,13 12 23% 4,47-5,79 0,056 4 frozen +UPD 7,04 12 34% 3,79-6,48 5 fresh +UPD 29,36 13 6% 28,38-30,34 0,0008 5 frozen +UPD 31,66 13 5% 30,73-32,58
9 0 10 20 30 40 50 60 Co nc. H NL ( ng /m l) Sample -UPD Sample 1 Sample 2 Sample 3 0 10 20 30 40 50 60 Co nc. H NL ( ng /m l) Sample +UPD Sample 1 Sample 2 Sample 3
buffer and a mean concentration of 60,97 ng/ml for the frozen samples treated with UPD-buffer.
There was a statistically significant difference in sample 5 (p=0,0008) and sample 6 (p=4,1E-07), with levels being higher in the frozen samples compared to the fresh samples. There was not a statistically significant difference in sample 4 (p=0,056).
Fig. 1. Scatter plot showing the distribution of HNL-concentrations of frozen samples +/- UPD. The plot
to the left describes the distribution of samples -UPD for samples 1, 2 and 3. The plot to the right describes the distribution of samples +UPD for samples 1, 2 and 3.
Fig. 2. Scatter plot showing the distribution of HNL-concentrations of fresh or frozen samples +UPD. The
plot to the left describes the distribution of fresh samples +UPD for samples 4, 5 and 6. The plot to the right describes the distribution of frozen samples +UPD for samples 4, 5 and 6.
10 EPX
The analysis of EPX was designed identically to the analysis of HNL - it was divided into two parts. One experiment compared frozen and thawed urine samples that were treated with UPD-buffer to frozen and thawed urine samples left untreated as a reference. The other experiment compared fresh samples to frozen and thawed samples where both were treated with UPD-buffer. It is important to note that the samples for EPX were collected from the first morning urine. All samples that were to be tested for EPX were correlated to the urinary creatinine (u-Crea) of the sample to eliminate the variable of the glomerular filtration rate of the individual. This makes the quantification more correct, as the water-intake of the individual is taken into consideration, which gives a more realistic view of the protein concentration in the sample:
• Sample 7: 10,82 mmol/L • Sample 8: 9,47 mmol/L • Sample 9: 11,82 mmol/L • Sample 10: 3,44 mmol/L • Sample 11: 8,34 mmol/L • Sample 12: 8,90 mmol/L
The results of the first experiment, using frozen and thawed urine, are shown in the upper part of table 2 and a scatter plot of the results can be seen in figure 3. The samples are numbered 7, 8 and 9. Sample 7 was run 15 times, giving a mean concentration of 32,80 μg/mmol Crea for the samples not treated with UPD-buffer and a mean concentration of 48,09 μg/mmol Crea for the samples treated with UPD-buffer. Sample 8 was run 14 times giving a mean concentration of 78,39 μg/mmol Crea for samples not treated with UPD and a mean concentration of 79,16 μg/mmol Crea for samples treated with UPD-buffer. Sample 9 was run 15 times giving a mean concentration of 28,85 μg/mmol Crea for samples not treated with UPD and a mean concentration of 31,56 μg/mmol Crea for samples treated with UPD-buffer. A reference value for the concentration of EPX in healthy individuals (established from 313 individuals) is around 61 μg/mmol Crea. (Gore et al. 2003).
Table 2. A summary of the statistics of EPX. Mean indicates the mean concentration of n number of
experiments. The CV% is calculated, as well as a 95% CI for each group and individual. The p-value (p<0.05 is considered significant) for each sample is also indicated as calculated by a student’s t-test.
There was a statistically significant difference between the treated samples and the untreated samples only in sample 7 (p=1,06E-07), with levels being higher in the samples treated with the UPD-buffer. Sample 8 (p=0,854) and sample 9 (p=0,055) showed no statistically significant
Sample Mean n CV% 95%CI p
7 -UPD 32,80 15 20% 29,00-35,56 1,06E-07 7 +UPD 48,09 15 10% 46,35-49,82 8 -UPD 78,39 14 16% 71,91-84,87 0,854 8 +UPD 79,16 15 12% 74,18-84,14 9 -UPD 28,85 15 15% 26,62-31,09 0,055 9 +UPD 31,56 15 13% 29,53-33,6 10 fresh +UPD 63,91 15 5% 61,28-66,54 0,399 10 frozen +UPD 62,99 15 8% 61,35-64,62 11 fresh +UPD 73,32 14 6% 71,23-75,41 0,016 11 frozen +UPD 68,64 14 5% 66,39-70,89
11 0,00 20,00 40,00 60,00 80,00 100,00 μ g E P X/m m o l Cre a Sample -UPD Sample 7 Sample 8 Sample 9 0,00 20,00 40,00 60,00 80,00 100,00 μ g E P X/m m o l Cre a Sample +UPD Sample 7 Sample 8 Sample 9 30,00 40,00 50,00 60,00 70,00 80,00 90,00 μ g E P X/m m o l Cre a Fresh +UPD Sample 10 Sample 11 Sample 12 30,00 40,00 50,00 60,00 70,00 80,00 90,00 μ g E P X/m m o l Cre a Frozen +UPD Sample 10 Sample 11 Sample 12 difference between the treated and the untreated samples.
The second experiment, comparing fresh samples to frozen samples with all samples treated with UPD-buffer, are shown in the lower part of table 2 and a scatter plot of the results can be seen in figure 4. The samples are numbered 10, 11 and 12. Sample 10 was run 15 times, giving a mean concentration of 63,91 μg/mmol Crea for the fresh samples treated with UPD-buffer and a mean concentration of 62,99 μg/mmol Crea for the frozen samples treated with UPD-buffer. Sample 11 was run 14 times giving a mean concentration of 73,32 μg/mmol Crea for the fresh samples treated with UPD-buffer and a mean concentration of 68,64 μg/mmol Crea for the frozen samples treated with UPD-buffer. Sample 12 was run 14 times giving a mean concentration of 67,98 μg/mmol Crea for the fresh samples treated with UPD-buffer and a mean concentration of 61,04 μg/mmol Crea for the frozen samples treated with UPD-buffer.
There was a statistically significant difference between the frozen and fresh samples treated with the UPD-buffer in both sample 11 (p=0,016) and sample 12 (p=4,09E-05), with levels being higher in the fresh samples compared to the frozen samples. Sample 10 (p=0,399) showed no statistically significant difference between the treated and the untreated samples.
Fig. 3. Scatter plot showing the distribution of EPX-concentrations of frozen samples +/- UPD. The plot to
the left describes the distribution of samples -UPD for samples 7, 8 and 9. The plot to the right describes the distribution of samples +UPD for samples 7, 8 and 9.
Fig. 4. Scatter plot showing the distribution of EPX-concentrations of fresh or frozen samples +UPD. The
12
Discussion
When looking at the results from these experiments it is evident that the UPD-buffer works for both HNL and EPX. When comparing the frozen and thawed samples to the fresh samples there was a statistically significant difference with protein levels being higher in the frozen and thawed samples in two out of three individuals for HNL. For EPX, the levels were higher in the fresh samples that had not been frozen, with a statistically significant difference for two out of three individuals. This marks an interesting difference between the two proteins.
It is clear that there are a lot of individual variations in whether or not the UPD-buffer makes a difference when analyzing the levels of protein in the urine samples. This is probably due to the fact that there are individual variations in how much precipitates are formed in the urine. The formation of precipitates depends for example on the pH of the urine, the protein composition in the urine and a difference in the leakage of cells. Another factor could be that the individual in question had an infection or allergy at the time that the sample was collected. For example, when looking at figure 2, sample 6, there was almost a 50% increase in HNL-concentration measured when the sample had been frozen and then thawed (please refer to table 1 for mean values and p-values). An explanation for this could be that this individual had a higher amount of neutrophils in the urine, and when the sample was frozen these cells lysated and released HNL. Seeing the individual variations in excretion of HNL, with concentrations ranging from 0,87-44,61 ng/ml it would have been useful to correlate for u-Crea in these samples also, as was done in the samples tested for EPX. There is not as much individual variation in the excretion of EPX as there is of HNL. But naturally, as EPX is elevated in allergic conditions and these samples were collected from volunteers - with no regard to their health status - during the pollen-season in Sweden there might be a slight elevation in the individuals that were allergic.
Analyzing EPX generally seems to be problematic since there is normally a secretion of EPX in the urine independent of pathogenic conditions, as well as an elevation with unspecific Th2-reactions such as an allergy to for example birch pollen. To explain the results of the protein levels being generally higher in the fresh samples compared to the frozen and thawed samples one have to consider the structure of EPX. EPX is a very basic protein; EPX will readily stick to a lot of things, such as other proteins, plastics, tubes, pipettes, etc. An important factor in this is the pH of the urine; a lower pH will make the protein less sticky while a higher pH will make the protein more prone to attach itself to for example plastics. The pH of urine is normally neutral, with small variations depending on for example the diet of the individual. As EPX is such a sticky protein large amounts of EPX could be lost in handling of the samples. Examples of this could be the transferring of the sample between tubes, dilution or treatment with UPD-buffer, which requires additional pipetting. It is therefore important to note that the quantification of EPX in frozen and thawed urine samples may not be as exact as in the fresh samples. Frozen samples require more handling which presents more opportunities for the protein to stick to other things. In the future it would therefore be important to establish methods which are consequent and takes this into consideration. Only then a dependable and correct analysis of the protein concentration in the sample can be performed.
13 dissolves precipitates that are there, and does not affect the samples that does not contain as much precipitates.
Considering the new data and the new questions that have arisen, another important aspect comes to mind; are the quantification of these proteins - done by us and others - fully reliable? This study was done with very few observations and it would be useful for us to redo these experiments with more individuals in each group/for each protein to be able to draw any definitive conclusions. As mentioned above, it is important to establish consequent methods when analyzing these kinds of proteins and samples. As EPX is such a sticky protein, you cannot know how much of the protein is lost when transferring to new tubes, diluting the samples etc.. It would be useful to establish normal values and methods that are universally applicable, with clear explanations on for example what antibodies were used. Regarding antibodies it has been seen that the detection of HNL is highly dependent on the type of antibody used. HNL can exist in either monomers or dimers, and different antibodies can detect either monomers, dimers or both (Cai et al. 2010). It is therefore important to know which antibody you use for your experiment, and which configuration of the protein you want to detect.
14
References
Arena, A., Stassi, G., Iannello, D., Gazzara, D., Calapai, M., Bisignano, C., Bolignano, D., Lacquaniti, A., & Buemi, M. (2010). Both IL-1beta and TNF-alpha regulate NGAL expression in polymorphonuclear granulocytes of chronic hemodialysis patients. Mediators Inflamm, 2010, 613937.
Bochner, B. S. (2000). Systemic activation of basophils and eosinophils: markers and consequences. J Allergy Clin Immunol, 106(5 Suppl), S292-302.
Bystrom, J., Amin, K., & Bishop-Bailey, D. (2011). Analysing the eosinophil cationic protein--a clue to the function of the eosinophil granulocyte. Respir Res, 12, 10.
Cai, L., Rubin, J., Han, W., Venge, P., & Xu, S. (2010). The origin of multiple molecular forms in urine of HNL/NGAL. Clin J Am Soc Nephrol, 5(12), 2229-2235.
Collins, P. D., Marleau, S., Griffiths-Johnson, D. A., Jose, P. J., & Williams, T. J. (1995). Cooperation between interleukin-5 and the chemokine eotaxin to induce eosinophil accumulation in vivo. J Exp Med, 182(4), 1169-1174.
Flo, T. H., Smith, K. D., Sato, S., Rodriguez, D. J., Holmes, M. A., Strong, R. K., Akira, S., & Aderem, A. (2004). Lipocalin 2 mediates an innate immune response to bacterial infection by sequestrating iron. Nature, 432(7019), 917-921.
Gore, C., Peterson, C. G., Kissen, P., Simpson, B. M., Lowe, L. A., Woodcock, A., & Custovic, A. (2003). Urinary eosinophilic protein X, atopy, and symptoms suggestive of allergic disease at 3 years of age. J Allergy Clin Immunol, 112(4), 702-708.
Hogan, S. P., Rosenberg, H. F., Moqbel, R., Phipps, S., Foster, P. S., Lacy, P., Kay, A. B., & Rothenberg, M. E. (2008). Eosinophils: biological properties and role in health and disease. Clin Exp Allergy, 38(5), 709-750.
Humbles, A. A., Lu, B., Friend, D. S., Okinaga, S., Lora, J., Al-Garawi, A., Martin, T. R., Gerard, N. P., & Gerard, C. (2002). The murine CCR3 receptor regulates both the role of eosinophils and mast cells in allergen-induced airway inflammation and hyperresponsiveness. Proc Natl Acad Sci U S A, 99(3), 1479-1484.
Kita, H. (2011). Eosinophils: multifaceted biological properties and roles in health and disease. Immunol Rev, 242(1), 161-177.
Kjeldsen, L., Johnsen, A. H., Sengelov, H., & Borregaard, N. (1993). Isolation and primary structure of NGAL, a novel protein associated with human neutrophil gelatinase. J Biol Chem, 268(14), 10425-10432.
Lawrence, M. B., & Springer, T. A. (1993). Neutrophils roll on E-selectin. J Immunol, 151(11), 6338-6346.
15 Lonnberg, M., Drevin, M., & Carlsson, J. (2008). Ultra-sensitive immunochromatographic assay for
quantitative determination of erythropoietin. J Immunol Methods, 339(2), 236-244.
Miike, S., & Kita, H. (2003). Human eosinophils are activated by cysteine proteases and release inflammatory mediators. J Allergy Clin Immunol, 111(4), 704-713.
Morioka, J., Kurosawa, M., Inamura, H., Nakagami, R., Mizushima, Y., Chihara, J., Yokoseki, T., Kitamura, S., Omura, Y., & Shibata, M. (2000). Development of a novel enzyme-linked immunosorbent assay for blood and urinary eosinophil-derived neurotoxin: a preliminary study in patients with bronchial asthma. Int Arch Allergy Immunol, 122(1), 49-57.
Peters, A. M. (1998). Just how big is the pulmonary granulocyte pool? Clin Sci (Lond), 94(1), 7-19. Sanderson, C. J. (1992). Interleukin-5, eosinophils, and disease. Blood, 79(12), 3101-3109.
Sundd, P., Pospieszalska, M. K., Cheung, L. S., Konstantopoulos, K., & Ley, K. (2011). Biomechanics of leukocyte rolling. Biorheology, 48(1), 1-35.
Tischendorf, F. W., Brattig, N. W., Lintzel, M., Buttner, D. W., Burchard, G. D., Bork, K., & Muller, M. (2000). Eosinophil granule proteins in serum and urine of patients with helminth infections and atopic dermatitis. Trop Med Int Health, 5(12), 898-905.
Venge, P. (2004). Monitoring the allergic inflammation. Allergy, 59(1), 26-32.
Witko-Sarsat, V., Rieu, P., Descamps-Latscha, B., Lesavre, P., & Halbwachs-Mecarelli, L. (2000). Neutrophils: molecules, functions and pathophysiological aspects. Lab Invest, 80(5), 617-653.
Wolthers, O. D., & Heuck, C. (2003). Circadian variations in serum eosinophil cationic protein, and serum and urine eosinophil protein X. Pediatr Allergy Immunol, 14(2), 130-133.
Xu, S. Y., Carlson, M., Engstrom, A., Garcia, R., Peterson, C. G., & Venge, P. (1994). Purification and characterization of a human neutrophil lipocalin (HNL) from the secondary granules of human neutrophils. Scand J Clin Lab Invest, 54(5), 365-376.
Xu, S. Y., Pauksen, K., & Venge, P. (1995). Serum measurements of human neutrophil lipocalin (HNL) discriminate between acute bacterial and viral infections. Scand J Clin Lab Invest, 55(2), 125-131.
